OpenClaw remembers for OpenClaw. Sekha remembers for your full workflow

Name: OpenClaw remembers for OpenClaw. Sekha remembers for your full workflow
Availability: InStock
Author: sekha-ai

by sekha-ai·Feb 24, 2026·2 points·1 comment

Visit Project View on HN

AI Analysis

●●SolidBig BrainSolve My Problem

Portable memory for multi-LLM workflows, but Mem0 and MemGPT already solve this.

Strengths

•Genuine architectural insight: memory as separate governance layer, not agent-embedded.
•Real interop stack (MCP/REST/SDK/LiteLLM) means true tool portability across Claude, OpenClaw, Gemini.
•Rust + SQLite + Chroma stack is solid engineering for durability and semantic search.

Weaknesses

•Mem0 and MemGPT already offer API-first memory; differentiation unclear beyond self-hosted vs cloud.
•Landing page narrative ("Your AI gaslights you") oversells the problem—context windows exist by design, not a UX failure.

Post Description

OpenClaw's built-in memory is excellent—for OpenClaw. Markdown files, semantic search, survives restarts.

But it stays in OpenClaw.

I built Sekha for when you need memory that travels: OpenClaw today, Claude Code tomorrow, Kimi 2.5 or Gemini the next day. Intelligent embedding-based retrieval, persistent storage, universal API.

The difference: OpenClaw: MEMORY.md files, internal only

Sekha: SQLite + Chroma embeddings, REST/MCP/SDKs, any LLM via LiteLLM/OpenRouter

Use case: OpenClaw explores a codebase, stores findings in Sekha via MCP. Next day, Claude Code reads the same context via SDK. Your analytics pipeline queries it via REST. Same memory, any tool, any model.

Others add memory to OpenClaw. Sekha frees your memory from OpenClaw.

Stack: Rust (fast), SQLite (durable), Chroma (search), LLM-Bridge for universal routing. AGPL, self-hosted.

GitHub: https://github.com/sekha-ai/sekha-controller | Site: https://sekha.dev

The question: What would you build if your AI memory worked with every tool, not just one?

Similar Projects

AI/ML●●Solid

Sekha – What if AI remembered 3 years of conversations, not 3 hours?

The core idea is simple and pragmatic: attach a persistent, SQLite-backed vector store to any model so conversations don't vanish after a single context window. The repo leans into portability (Rust, self-hosted, AGPL) and the UI shows sensible controls like conversation folders and a context-budget token slider — useful details that suggest this is built for real use rather than a demo. My worry: retrieval quality, scaling and access controls will be the real battleground, not the clean chat UI.

Niche GemSolve My Problem

sekha-ai

113mo ago

AI/ML●●Solid